If you are unable to create a new account, please email support@bspsoftware.com

 

News:

MetaManager - Administrative Tools for IBM Cognos
Pricing starting at $2,100
Download Now    Learn More

Main Menu

Contributor users having difficulty saving elist items

Started by kevin.floyd, 09 Oct 2013 01:36:56 PM

Previous topic - Next topic

kevin.floyd

We have a budget model with close to 200 elist nodes.  Most users aren't having any problems, but a handful (5-6) have had trouble saving.  When they make a change and hit enter, it takes close 30 minutes to calc.  If you then try to save, the application hangs and never saves.  All other users have had no problems at all.  That leads me to believe it isn't a calculation or link issue.

Has anyone seen anything like this?  We also operate in a Citrix environment.  Are there issues with Citrix that could the trouble.  We are on Planning 10.1.1 without any fix packs.

kevin.floyd

Still running into the same issue and I'm at my wits end.  You key in a number and 12 minutes later you see the calculation.  You can't navigate in the application for another 15-20 minutes.  Could this be caused by something in the database?  I know you hit the database to save, but does Contributor interact with the database before the save?

topedgemonk


Though it still generates the question of why others users are not facing the same problem if that is the issue...

maybe circular references in d-links that get triggered when you save? Have you checked for those?

ericlfg

I would also possibly look into the cache files that are created for planning. 

In user's profile path (on citrix):
onlinecache101.at
onlinecache101.store

usually in the windows\temp path (on citrix):
jcecache101.at

Try moving these files onto the desktop of the citrix server (which will force the product to recreate them)

Also, some follow up questions:

1. Are specific elist items in a specific application affected for the users that are experiencing the issue?
2. Are specific cubes within the elist items affected, or all cubes?
3. On the citrix server when one of the affected users tries to save, is there a spike in the contributor.exe CPU usage in task manager? (associated to the user name)  How does this compare to other users that aren't affected?
4. Do you see any issues with elist performance during the load (initial open)?  Or just calculating / saving?

You may also want to get IBM support involved to setup a testcase and validate the behavior you're seeing.  If they are able to replicate, they might be able to get development assistance.

kevin.floyd

#4
Thanks, I will check into the suggestions.

Yes, IBM Support is involved and I believe as of last night finally was able to duplicate the errors.

It appears related to specific cubes and I am currently checking for circular d-link references.  And it relates to specific elist items in a specific application.  There appears to be no issue with performance during the load.

We did monitor task manager and didn't notice any spikes.  At least that's what IT said.  It just looked like a bunch of lines to me.


kevin.floyd

Just wanted to follow up.  Apparently, there was a problem with the links.  We had links simultaneously targeting formulas with one number and the detail of the formulas with another.  Because that particular cube used the P&L as a source, every cube we touched that fed the P&L created the problem.

ericlfg